X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fviki.py;h=3a7e9a0f2b42accd15d9589f7fbffd7f7b775562;hb=c59b61c0da8fc0fb73d745de59122e290a33b122;hp=234649ca8f8e885f488bbe976f57d7a487e60eef;hpb=1a83c731bd58ed85f6f7695cee9c88d09a224bc2;p=youtube-dl diff --git a/youtube_dl/extractor/viki.py b/youtube_dl/extractor/viki.py index 234649ca8..3a7e9a0f2 100644 --- a/youtube_dl/extractor/viki.py +++ b/youtube_dl/extractor/viki.py @@ -1,9 +1,11 @@ +# coding: utf-8 from __future__ import unicode_literals -import re +import json import time import hmac import hashlib +import itertools from ..utils import ( ExtractorError, @@ -11,10 +13,12 @@ from ..utils import ( parse_age_limit, parse_iso8601, ) +from ..compat import compat_urllib_request from .common import InfoExtractor class VikiBaseIE(InfoExtractor): + _VALID_URL_BASE = r'https?://(?:www\.)?viki\.(?:com|net|mx|jp|fr)/' _API_QUERY_TEMPLATE = '/v4/%sapp=%s&t=%s&site=www.viki.com' _API_URL_TEMPLATE = 'http://api.viki.io%s&sig=%s' @@ -22,27 +26,35 @@ class VikiBaseIE(InfoExtractor): _APP_VERSION = '2.2.5.1428709186' _APP_SECRET = '-$iJ}@p7!G@SyU/je1bEyWg}upLu-6V6-Lg9VD(]siH,r.,m-r|ulZ,U4LC/SeR)' - def _prepare_call(self, path, timestamp=None): + _NETRC_MACHINE = 'viki' + + _token = None + + def _prepare_call(self, path, timestamp=None, post_data=None): path += '?' if '?' not in path else '&' if not timestamp: timestamp = int(time.time()) query = self._API_QUERY_TEMPLATE % (path, self._APP, timestamp) + if self._token: + query += '&token=%s' % self._token sig = hmac.new( self._APP_SECRET.encode('ascii'), query.encode('ascii'), hashlib.sha1 ).hexdigest() - return self._API_URL_TEMPLATE % (query, sig) + url = self._API_URL_TEMPLATE % (query, sig) + return compat_urllib_request.Request( + url, json.dumps(post_data).encode('utf-8')) if post_data else url - def _call_api(self, path, video_id, note, timestamp=None): + def _call_api(self, path, video_id, note, timestamp=None, post_data=None): resp = self._download_json( - self._prepare_call(path, timestamp), video_id, note) + self._prepare_call(path, timestamp, post_data), video_id, note) error = resp.get('error') if error: if error == 'invalid timestamp': resp = self._download_json( - self._prepare_call(path, int(resp['current_timestamp'])), + self._prepare_call(path, int(resp['current_timestamp']), post_data), video_id, '%s (retry)' % note) error = resp.get('error') if error: @@ -55,10 +67,31 @@ class VikiBaseIE(InfoExtractor): '%s returned error: %s' % (self.IE_NAME, error), expected=True) + def _real_initialize(self): + self._login() + + def _login(self): + (username, password) = self._get_login_info() + if username is None: + return + + login_form = { + 'login_id': username, + 'password': password, + } + + login = self._call_api( + 'sessions.json', None, + 'Logging in as %s' % username, post_data=login_form) + + self._token = login.get('token') + if not self._token: + self.report_warning('Unable to get session token, login has probably failed') + class VikiIE(VikiBaseIE): IE_NAME = 'viki' - _VALID_URL = r'https?://(?:www\.)?viki\.com/(?:videos|player)/(?P[0-9]+v)' + _VALID_URL = r'%s(?:videos|player)/(?P[0-9]+v)' % VikiBaseIE._VALID_URL_BASE _TESTS = [{ 'url': 'http://www.viki.com/videos/1023585v-heirs-episode-14', 'info_dict': { @@ -120,6 +153,23 @@ class VikiIE(VikiBaseIE): 'like_count': int, 'age_limit': 13, } + }, { + # youtube external + 'url': 'http://www.viki.com/videos/50562v-poor-nastya-complete-episode-1', + 'md5': '216d1afdc0c64d1febc1e9f2bd4b864b', + 'info_dict': { + 'id': '50562v', + 'ext': 'mp4', + 'title': 'Poor Nastya [COMPLETE] - Episode 1', + 'description': '', + 'duration': 607, + 'timestamp': 1274949505, + 'upload_date': '20101213', + 'uploader': 'ad14065n', + 'uploader_id': 'ad14065n', + 'like_count': int, + 'age_limit': 13, + } }, { 'url': 'http://www.viki.com/player/44699v', 'only_matching': True, @@ -128,26 +178,6 @@ class VikiIE(VikiBaseIE): def _real_extract(self, url): video_id = self._match_id(url) - streams = self._call_api( - 'videos/%s/streams.json' % video_id, video_id, - 'Downloading video streams JSON') - - formats = [] - for format_id, stream_dict in streams.items(): - height = self._search_regex( - r'^(\d+)[pP]$', format_id, 'height', default=None) - for protocol, format_dict in stream_dict.items(): - if format_id == 'm3u8': - formats = self._extract_m3u8_formats( - format_dict['url'], video_id, 'mp4', m3u8_id='m3u8-%s' % protocol) - else: - formats.append({ - 'url': format_dict['url'], - 'format_id': '%s-%s' % (format_id, protocol), - 'height': height, - }) - self._sort_formats(formats) - video = self._call_api( 'videos/%s.json' % video_id, video_id, 'Downloading video JSON') @@ -159,7 +189,7 @@ class VikiIE(VikiBaseIE): title = 'Episode %d' % video.get('number') if video.get('type') == 'episode' else video.get('id') or video_id container_titles = video.get('container', {}).get('titles') if container_titles: - container_title = container_titles.get('en') or container_titles[titles.keys()[0]] + container_title = container_titles.get('en') or container_titles[container_titles.keys()[0]] title = '%s - %s' % (container_title, title) descriptions = video.get('descriptions') @@ -186,7 +216,7 @@ class VikiIE(VikiBaseIE): 'videos/%s/subtitles/%s.%s' % (video_id, subtitle_lang, subtitles_format)), } for subtitles_format in ('srt', 'vtt')] - return { + result = { 'id': video_id, 'title': title, 'description': description, @@ -196,14 +226,43 @@ class VikiIE(VikiBaseIE): 'like_count': like_count, 'age_limit': age_limit, 'thumbnails': thumbnails, - 'formats': formats, 'subtitles': subtitles, } + streams = self._call_api( + 'videos/%s/streams.json' % video_id, video_id, + 'Downloading video streams JSON') + + if 'external' in streams: + result.update({ + '_type': 'url_transparent', + 'url': streams['external']['url'], + }) + return result + + formats = [] + for format_id, stream_dict in streams.items(): + height = int_or_none(self._search_regex( + r'^(\d+)[pP]$', format_id, 'height', default=None)) + for protocol, format_dict in stream_dict.items(): + if format_id == 'm3u8': + formats = self._extract_m3u8_formats( + format_dict['url'], video_id, 'mp4', m3u8_id='m3u8-%s' % protocol) + else: + formats.append({ + 'url': format_dict['url'], + 'format_id': '%s-%s' % (format_id, protocol), + 'height': height, + }) + self._sort_formats(formats) + + result['formats'] = formats + return result + -class VikiChannelIE(InfoExtractor): +class VikiChannelIE(VikiBaseIE): IE_NAME = 'viki:channel' - _VALID_URL = r'https?://(?:www\.)?viki\.com/tv/(?P[0-9]+c)' + _VALID_URL = r'%s(?:tv|news|movies|artists)/(?P[0-9]+c)' % VikiBaseIE._VALID_URL_BASE _TESTS = [{ 'url': 'http://www.viki.com/tv/50c-boys-over-flowers', 'info_dict': { @@ -220,17 +279,25 @@ class VikiChannelIE(InfoExtractor): 'description': 'md5:05bf5471385aa8b21c18ad450e350525', }, 'playlist_count': 127, + }, { + 'url': 'http://www.viki.com/news/24569c-showbiz-korea', + 'only_matching': True, + }, { + 'url': 'http://www.viki.com/movies/22047c-pride-and-prejudice-2005', + 'only_matching': True, + }, { + 'url': 'http://www.viki.com/artists/2141c-shinee', + 'only_matching': True, }] - _API_BASE = 'http://api.viki.io/v4/containers' - _APP = '100000a' + _PER_PAGE = 25 def _real_extract(self, url): channel_id = self._match_id(url) - channel = self._download_json( - '%s/%s.json?app=%s' % (self._API_BASE, channel_id, self._APP), - channel_id, 'Downloading channel JSON') + channel = self._call_api( + 'containers/%s.json' % channel_id, channel_id, + 'Downloading channel JSON') titles = channel['titles'] title = titles.get('en') or titles[titles.keys()[0]] @@ -239,17 +306,17 @@ class VikiChannelIE(InfoExtractor): description = descriptions.get('en') or descriptions[descriptions.keys()[0]] entries = [] - for video_type in ('episodes', 'clips'): - page_url = '%s/%s/%s.json?app=%s&per_page=%d&sort=number&direction=asc&with_paging=true&page=1' % (self._API_BASE, channel_id, video_type, self._APP, self._PER_PAGE) - while page_url: - page = self._download_json( - page_url, channel_id, - 'Downloading %s JSON page #%s' - % (video_type, re.search(r'[?&]page=([0-9]+)', page_url).group(1))) + for video_type in ('episodes', 'clips', 'movies'): + for page_num in itertools.count(1): + page = self._call_api( + 'containers/%s/%s.json?per_page=%d&sort=number&direction=asc&with_paging=true&page=%d' + % (channel_id, video_type, self._PER_PAGE, page_num), channel_id, + 'Downloading %s JSON page #%d' % (video_type, page_num)) for video in page['response']: video_id = video['id'] entries.append(self.url_result( - 'http://www.viki.com/videos/%s' % video_id, 'Viki', video_id)) - page_url = page['pagination']['next'] + 'http://www.viki.com/videos/%s' % video_id, 'Viki')) + if not page['pagination']['next']: + break return self.playlist_result(entries, channel_id, title, description)